[U-Boot] [PATCH v3 0/7] Fixes/Addition to use the USB Ethernet gadget with the DWC3 gadget controller

[Eugeniu Rosca]

On Wed, Aug 8, 2018 at 6:54 PM Tom Rini <trini at konsulko.com> wrote:
>
> On Wed, Aug 08, 2018 at 06:44:07PM +0200, Eugeniu Rosca wrote:
> > On Wed, Aug 8, 2018 at 6:18 PM Jean-Jacques Hiblot <jjhiblot at ti.com> wrote:
> > > On 08/08/2018 15:08, Eugeniu Rosca wrote:
> > [snip]
> > > > I think having a warning-free build is a basic policy everybody is
> > > > expected to comply with.
> > > I agree. I meant that it could be fixed at the time of the commit by the
> > > maintainer.
> >
> > I'm not sure how well this plays with https://developercertificate.org/.
> > At least for my contributions, I would expect that they reach the main
> > tree unmodified, unless there is some trivial change in commit
> > metadata (which I expect to be documented). But I'll let others to
> > comment, who went through such kind of questions before.
>
> In DCO terms, it is my belief that a maintainer can make additional
> changes, so long as they too add a S-o-B line.  I also like it if they
> do:
> [me: Did stuff]
> so it's clear what changed.  But I don't think that's a conflict with
> DCOs.

One interesting aspect is whether the community has a chance to review
the changes occurred in the timeframe between the author's and the
maintainer's S-o-B. This seems to be elided in DCO, probably because
DCO treats all contributors equally. But obviously, the maintainer has
an outstanding status of committing the work into the tree.

It is my assumption that the maintainer should give enough time to
community to reach some agreement that the contents is ready for
merge. With this assumption in mind, it should probably be an
exception for the maintainer to perform any changes without prior
community review. This is why I noted that I expect a v4 to fix the
compiler warning (especially given that this is not a metadata
change). But don't get me wrong. My main interest is fixing the gcc
warning occurring in sandbox build with this patch-set. I am open
minded about the rest.

> In custodian terms, it's a whole lot quicker if someone can directly
> apply a series rather than apply, rebase (fixup a patch) and continue.
> So, if you do another round and fix the problem it'll be one less thing
> for the custodian to deal with.
>
> --
> Tom

Thanks,
Eugeniu.